AI Job Search: The Job Search That Runs on Your Machine
An open-source job-hunting framework that runs on your own machine: find postings, tailor your CV and cover letter, and prep interviews with Claude Code.
Project intro
Looking for work means writing a fresh, tailored CV and cover letter for every role, then keeping track of dozens of applications, interviews and follow-ups — slow, repetitive work that is easy to lose track of. AI Job Search turns that whole chain into one workflow that runs on your own machine: find postings, judge the fit, tailor the application, prep for the interview, and track the result. It is free and open source, and more than 71,000 people have starred it on GitHub.
Six slash commands carry you through it. /setup builds your profile from your documents or a past CV; /scrape searches job boards for matching postings; /rank shortlists the best fits; /apply drafts a tailored CV and cover letter; /interview builds a prep pack and runs a mock interview; /outcome records what happened and drafts follow-ups. When you apply, one agent drafts and a second agent, starting fresh, checks the drafts against the company — then everything is compiled into real PDFs, a two-page CV and a one-page cover letter, and checked the way a hiring system's parser would read them.
The author built it for his own job search after losing his geophysicist job in late 2025, and used it week after week on his own career: 69 tailored applications led to 20 first interviews and one signed contract, and he started as an AI engineer in June 2026. Your data stays in your own files on your own machine, and nothing in the CV is invented — every claim is checked against your actual profile.
Two honest notes before you start. You need a Claude Code subscription or API credits, plus Python 3.10+, Bun and LaTeX, so this is a setup job, not a one-click app. And the built-in job board search is made for Danish boards such as Jobindex, so outside Denmark you will need its /add-portal command to wire up your local job boards. It is MIT-licensed, with fresh code landing on 12 November 2026.
Actions
Tech stack
Project info
- Project category
- Productivity
- Project status
- Online
- Collaboration intent
- None